Statistical methods for analysis of in vitro antigen quantification data for veterinary biologic products.

In vitro assay systems that measure the antigen content of veterinary biologic products use a statistical model for quantification. The most prevalent assay system for measuring veterinary biologics in the United States compares a reference, demonstrated to be protective in the host species, to the serial being tested. An ideal statistical model evaluates the similarity of response and compares the amount of antigen of a serial to the reference. This method is defined as the relative potency approach. Various statistical methods have been proposed for relative potency determinations. The standard curve, slope ratio assay, parallel line assay, four-parameter (logistics) fit, and integration of the area under the curve have been used. Each statistical method has strengths and weaknesses. The parallel line assay accepts data from assays with non-ideal dose response curves, provides a similarity comparison, and evaluates the relative potency in the most linear portion of the curve. The parallel line method is recommended for evaluation of in vitro assays for potency determination of U.S. licensed veterinary biologics.